A. K. Narain University of Wisconsin, Madison, USA

Heinz Bechert Universitat Gottingen, FRG Lewis Lancaster University of California, Berkeley, USA B.J. Stavisky WN1IR, Moscow, USSR Leon Hurvitz UBC, Vancouver, Canada Alexander W. MacDonald Universite de Paris X, Nanterre, France Alex Wayman Columbia University, New York, USA

Stephan Beyer University of Wisconsin, Madison, USA

Volume 3


Number 1


I. ARTICLES 1. A Study of the Madhyamika Method of Refutation, Especially of its Affinity to that of Kathdvatthu, by Shohei Ichimura Prajnaparamita and the Buddhahood of the Non-Sentient World: The San-Lun Assimilation of Buddha-Nature and Middle Path Doctrine, by Aaron K. Koseki A Clue to the Authorship of the Awakening of Faith: "&k$ananda's" Redaction to the Word "Nien," by Whalen W. Lai The Abhidharmika Notion of Vijndna and its Soteriological Significance, by Braj M. Sinha Some Comments on Tsong kha pa's Lam rim chen mo and Professor Wayman's Calming the Mind and Discerning the Real, by Geshe Sopa Alex Wayman Replies to Geshe Sopa Geshe Sopa Replies to Alex Wayman





34 54

4. 5.

68 93 98

II. SHORT PAPERS 1. 2. Archaeological Excavations at Piprahwa and Ganwaria and the Identification of Kapilavastu, by K. M. Srivastava 103 Notes on the Textcritical Editing of the Bodhisattvdvaddnakalpalatd, by Frances Wilson 111

III. BOOK REVIEWS AND NOTICES 1. Santideva: Mystique bouddhiste des Vile et VHIe siecles, by Amalia Pezzali 115 2. On Knowing Reality: The Tattvartha Chapter of Asariga's Bodhisattvabhumi, by Janice Dean Willis 117 3. Women in Buddhism: Images of the Feminine in Mahayana Tradition, by Diana Y. Paul 120 4. Wittgenstein and Buddhism, by Chris Gudmunsen. 122

IV. NOTES AND NEWS 1. A Report on the 2nd Conference of the IABS 2. Report on A Critical Dictionary of Pali 127 130



A Study of the Madhyamika Method of Refutation, Especially of its Affinity to that of Kathdvatthu

by Shohei Ichimura

-IIt was a quarter of a century ago that Prof. T.R.V. Murti published his work on the Madhyamika philosophy,1 which has made one essential point of Madhyamika negation thoroughly known to post-war scholarship, i.e., that Madhyamika philosophy is a critique of all philosophical theses, and that this critique does not imply any thesis of its own but exclusively reveals an inherent self-contradiction in any and every philosophical thesis. This method has been called reductio-ad-absurdum argument after the fashion Stcherbatsky used. My use of the term Madhyamika Dialectic is also in this particular sense. Although Madhyamika dialectic is an age-old subject and has also been treated in modern scholarship frequendy, I found that little attention has been given to the fact that there is an intrinsic affinity between the Madhyamika and the pre-classical Abhidharmist methods of refutation. I am especially referring to the method which is recorded in the Kathdvatthu or the Points of Controversy.2 To demonstrate this affinity is my primary objective in this paper. In order to accomplish this demonstration, first, I will try to show why the pre-classical Buddhist debators knew the two basic rules of Syllogistic Inference, namely anvaya and vyatireha, which I may translate as 'positive and contrapositive instantiations.' They applied these rules in order to defend their own thesis in terms of logicality, while refuting the other's in terms of illogicality. Secondly, I shall make some point of affinity clear as to the Madhyamika method in parallel to that of the Kathdvatthu. 7

-IIIn the Vigrahavydvartani, and especially in his self-commentary, Nagarjuna frequently equates the term nihsvabhava, or absence of ownbeing with that of sunyatd, or emptiness.3 He states, for instance, that 'light' and 'darkness' do not possess their own-being. They have no selfidentifying essence, because they are co-relative, mutually interdependent, and hence, unable to come into existence by themselves. Traditionally, this absence of own-being, as equated by Nagarjuna with the concept of sunyatd, has been regarded as one of the hardest subjects for rational and intellectual understanding, because the method of its exposition was and has been exclusively dialectical. In common sense thinking, we regard both the faculty of cognition (pramdna) and its object iprameya) as two different things, and yet we take their coalescence for granted for the fact of cognition. This is comparable to the fact of illumination in which both a candle light and its object of illumination somehow partake. As the major objective of the Vigrahavydvartani, Nagarjuna applies his dialectic to this particular context. I found an intrinsic affinity between his dialectical method as applied in this text and that of the Kathdvatthu. We know that our language works not only as an instrument for expressing inner thoughts, but also as an instrument for their communication. Because of this, we accept that each and every word or sentence has its 'own-being,' or its self-identifying principle. For, such own-being constitutes not only the essence of a conception that each expression signifies, but also of an external existent it refers to. Accordingly, the above two terms, i.e., light and darkness, are regarded as differentiating their respective meanings in our consciousness as well as their respective objects extraneous to us as referents. Nevertheless, Nagarjuna and his Madhyamika followers state that our very insistence upon the nature and function of our language convention ipso facto commits us to the truth of the reciprocal dependence of any two related terms and sentences, and that we are in fact subscribing ourselves to the truth of absence of own-being.4 The question is: How and why can the Madhyamika dialectic be said to be the only method that is capable of revealing the fact of universal reciprocity (paraspardpeksata) and absence of ownbeing {nih-svabhdvatd}} I believe that the aspect of universal interdependence can be disclosed by logically analysing the Madhyamika method of refutation.


— Ill— In Indian Syllogistic Inference, when two predications are related causally or tautologically, they constitute a necessary relation which Buddhist and Hindu logicians called vydpti. Logically, this relation or pervasion of one predication by another consists of the preceding predication of Reason (hetu) and of the subsequent one as Thesis or Conclusion (sddhya). As briefly referred to, Dignaga (a Buddhist logician of the 5th century) introduced three conditions for the validity of any given vydpti and theorized the dual rules of positive and contrapositive instantiations. In order to explain these rules, let me take one of the stock examples of the classical logicians, namely an inference of the existence of fire from the existence of smoke. In order for a person to let others know a breakout of a fire on a distant mountain, what he must do is to remind the listener of the commonly-known concomitance between smoke and fire by pointing at the rising smoke. On the part of the listener, upon perceiving a particular smoke rising from the slope of the mountain, his mind regulates itself into recalling a causal association of smoke and fire. // is this logical process of the mind itselfthat actually consists of the dual instantiations positive and contrapositive. I shall explain this point briefly. In order than an inference be correct, two logically related predications (hetu and sddhya) must be verified by the substratum intended to be known. In our example the substratum, such as a mountain, must be able to bear smoke as well as fire simultaneously, because, otherwise, it may happen that the mountain may bear smoke but not fire, and vice versa. If this is the case with the substrarum in question, it is obvious that no valid inference becomes possible. The speaker, therefore, is obliged to demonstrate whether the substratum (a mountain) is a member of the class whose members are invariably able to bear smoke and fire. This is called positive instantiation, or anvaya. The same speaker, however, is also obliged to give contrapositive instantiation as to whether the substratum in question is clearly outside the scope of the contrapositive class, because the latter members are neither capable of bearing fire nor smoke. This is contrapositive instantiation, or vyatireka, for which the speaker presents an instance, such as a lake or a water-dam, etc., where neither of the two predications can be applied. In short, by means of dual processes of instantiations, the speaker can confirm the demarcation between posidve and contrapositive classes (sapaksa and vipaksa resp.) and thereby determine the particular substratum as a member of the positive one. 5 9

As to the question of why appeal has to be made not only to positive instantiation but also to the contrapositive one, I believe it will become self-evident in my subsequent demonstration. Here it suffices to say that positive instantiation alone cannot fully differentiate those class members which are either "capable of bearing smoke but not fire," or "capable of bearing fire but not smoke." Suppose when, knowingall this, someone encounters an opponent in the arena of debate. How should he conduct his argument? He has to explore every possible error in his opponent's logic. Sometimes, he may even try to deliberately induce logical errors in his opponent. Nevertheless, he is obliged to abide in accordance with the basic rules of logic, such as dual instantiations. I believe that the debators of the Kathdvatthu applied such method and in following their step Nagarjuna innovated his Madhyamika method of refutation.

-IVThere is clear evidence for the fact that the pre-classical Buddhist debators were fully aware of the dual rules of logical instantiation, and applied these as a method of refutation. The procedure of argument in the text is so repetitive in form that I shall have to take up only the initial refutation. The controversy here is concerned with the status of pudgala. The orthodox Theravadin who rejects the reality of pudgala faces the challenge by the Pudgalavadin in the arena of debate. Formally, the refutation consists offiveconsecutive sessions. First, the Theravadin presents (1) Refutation against the Pudgalavadin, which is followed by the latter's (2) Rejoinder, (3) Refutation, (4) Application, and (5) Conclusion.*' Their arguments invariably include the dual demonstrations being applied positively and contrapositively.

No. I Two related predications cum substratum "P": (Pudgala) "is known in the sense of a genuinely real thing" (puggalo upalabbhati saccikatthaparamatthendti) "Q": (Pudgala) "is known in the same way a genuinely real thing is known" (yo saccikattho paramattho tato so puggalo upalabbhati saccikatthaparamatthendti) 10

T h e two predications, which I quoted from Mrs. Rhys Davids' translation, can be transcribed as "P" and "Q" respectively. A notation "P" is given to the predication: kpudgala "is known in the sense of a genuinely real thing," and another notation "Q" to: A pudgala "is known in the same way a genuinely real thing is known."

No. 2 Positive (anvaya) and Contrapositive (vyatireka) Instantiations: Theravadin: "P and Q" is assumed to be verified by substratum (sapaksa) such as dharmas, while "-Q and -P" is to be falsified by substratum (vipaksa) such as pudgalas. Pudgalavadin: "Q and P" is assumed to be verified by substratum, such as pudgalas as well as dharmas, while "-P and -Q" is to be verified by all those remaining. In No. 2,1 specify two mutually contrary concomitances as well as their respective contrapositions, which the Theravadin and the Pudgalavadin apply throughout their demonstration. It is also intended to show the workings of positive and contrapositive instantiations which respectively determine the positive and contrapositive classes (sapaksa and vipaksa). For the Theravadin, dharmas alone are real, and hence, they constitute the positive class. Accordingly, the position "P and Q" and its contraposition "-Q and -P" should respectively serve as criteria to distinguish whatever is real like a dharma (sapaksa) and whatever is unreal like an empirical person (vipaksa). For the Pudgalavadin, however, applying the same concomitance as that of the Theravadin is obviously disadvantageous. Therefore, he introduces an exacdy contrary concomitance to refute the Theravadin, namely, "Q and P" and its contraposition, "-P and -Q". But the Theravadin logical strategy, and especially the Pudgalavadin's, cannot be understood fully without help from the Western form of logical implication, which I prepare in No. 3.

No. 3 Hypothetical Syllogism based upon "P then Q" and "Q then P": If "P then Q," and "P," therefore "Q." (modus ponendo ponens) 11

If "P then Q," and "-Q," therefore "-P." (modusponendo tollens) If "Q then P," and "Q," therefore "P." If "Q then P," and "-P," therefore "-Q." T h e one obvious reason for the usefulness of western forms is evident in the chart given in No. 4, i.e., the antecedent statement always binds to the consequent, and this conditional implication can best be expressed in the hypothetical syllogism of the West. No. 4 Five Refutational Sessions between Theravadin and Pudgalavadin: I Theravadin Refutation Pudgalavadin thesis "P.-Q" is false, because P D Q ; "P.-Q" is false, because -Q = > -P; Therefore, Pudgalavadin thesis "P.-Q" is false. Ill Pudgalavadin Refutation Theravadin thesis "-P.Q" can be refuted, because -P ^ - Q ; "-P.Q" can be refuted, because Q 3 P; Therefore, Theravadin thesis "-P.Q" can be refuted. II Pudgalavadin Rejoinder Theravadin thesis "-P.Q" is false, because -P = > -Q; "-P.Q" is false, because Q = > P; Therefore, Theravadin thesis "-P.Q" is false. IV Pudgalavadin Application Our thesis "P.-Q" is not falsified, and Your refutation "-(P.-Q)" is not acceptable, because P o Q, and -Q = >P Therefore, your refutation "-(P.-Q)" is not acceptable.

Pudgalavadin Conclusion Our thesis "P.-Q" is not refuted, because "P.Q" is not compelled; Your refutation "-(P.-Q)" is not convincing, because "-Q.-P" is not compelled; Because "P.Q" and "-Q.-P" are not compelled, our thesis "P.-Q" is not refuted. No. 4 shows my transcription of the five consecutive sessions of arguments. 6 I consistently replace the form of Indian logical concomitance with that of Western logical implication. As a result, the chart shows not only the dynamism of the sessions but also the logical context in which 12

indeterminancy ensued. There seem to be two basic reasons for the indetermination of the controversy: (1) both parties violated the logical boundary of positive and contrapositive classes, and (2) this in turn allowed the Pudgalavadin to apply the contrary implication. First, for the Theravadin, both predications "P" and "Q" should be verified by the substratum of dharmas, i.e., a dharma "is known in the sense of a genuinely real thing" (="P"), and "is known in the same way a genuinely real thing is known" (="Q"). Therefore, he uses this implication as a criterion to defend the reality of dharmas and to refute the Pudgalavadin heresy thatpudgalas are also real. But he faces a problem here, because, he cannot reject "Q" about pudgalas though no problem to d o so with "P." This means that the Theravadin violated the logical boundary ofsapaksa and vipaksa in applying "Q" not only to dharmas but also to pudgalas. This logical ambivalence is in fact derived from doctrinal reasons. T h e repudiation of an empirical person (pudgala) constitutes the core of Buddhist doctrine. T h e Theravadin is obliged to assert "Q" because the unreality of pudgalas is knowable only through the way the reality of psycho-physical elements (dharmas) is known. To further complicate the matter, the Pudgalavadin also shows a similar logical ambivalence due to similar doctrinal reasons. He asserts "P" about pudgalas but fails to assert "Q," because if he does so, he is ipso facto completely identifying pudgala with dharma, which is heresy for the Pudgalavadin as well. Logically, he also violates the boundary between the classes of dharmas arid pudgalas in applying "P" to them equally. Second, the strike of ingenuity on the part of the Pudgalavadin is the use of contrary implication as a weapon to demonstrate the logical vulnerability of his opponent. This possibility must have been intuited from the fact that the two contestants stood in an exact contrariety, i.e., "-P.Q" by the Theravadin and "P.-Q" by the Pudgalavadin. Their forces of argument, as shown in the chart, come to an equal balance. The Theravadin argues: If you Pudgalavadin assert the reality of pudgala "P," you are also obliged to assert its knowability in the same way as dharmas are known "Q." But you do not, i.e., "-Q." If you do not assert "Q," you are also obliged logically not to assert "P," i.e., "-Q 3 -P." But you assert "P." Therefore, your claim "P.-Q" is false. Now, the Pudgalavadin replies: If you Theravadin assert the knowability of pudgala in the same way dharmas are known "Q," you should also assert its reality "P." But you do not, i.e., "-P." Since you do not assert "P," you are also obliged logically not to assert 4 Q," i.e., " - P ^ - Q . " But you assert "Q." Therefore, your claim "-P.Q" is false. 13

-VMy finding about the Madhyamika method as parallel to that of the Kathdvatthu is rather a simple one, namely that it seeks to create a disconnective relationship between conceptual terms, predications, or propositions in the forms of "P.-Q" and "-P.Q." If one speaks of a motion, for instance, we can match his statement with another about its agent as regards to their relationship. 7 In the Vigrahavydvartani, Nagarjuna creates this particular context by the metaphor of'light' and 'darkness.' H e reminds us that our cognition always involves cognizer and cognized just as the fact of illumination. He assigns predications to the ilumining and the illumined respectively as "is capable of illumining" ("P") and "is capable of darkening" ("Q"). By applying both predications to 'light' and 'darkness,' he obtains the formulas of "P.-Q" and "-P.Q." He argues: Wherever there is a light illumining, there should be no darkness ("P.Q") and vice versa ("-P.Q"), which means that the two never can meet. 8 T h e state of affairs is precisely parallel to that of the Kathdvatthu controversy. How did Nagarjuna try to solve this logical absurdity? As I understand, he generally takes two approaches. First, in accordance with convention, which assumes both 'light' and 'darkness' for the fact of illumination, he points out that the only way to make this positive concomitance "P.Q" possible is to repudiate the concept of own-being (self-identifying principle) from these entities, so as to accept light and darkness in terms of their reciprocal exchangeability. Second, in accordance to trans-convention, he repudiates both "P.-Q" and "-P.Q," which he must have justified in reference to two contrary implications "P o Q" and "Q 3 P" in parallel to the Kathdvatthu controversy. Here may I point out the fact that "P D Q" and "Q ^ P" together express logical reciprocity. In concluding my paper, I am obliged to state two points: (I) As evident in my demonstration, the logical concomitance of two predications differentiated four different classes of variables. I believe that the Buddhist fourfold logical category such as catuskoti has its relevant basis in this logic of concomitance. (2) Our conceptual and logical treatment of religious insight in general has its own limitation. T o deal with the statements that refer to the dialectical dimension which bridges the empirical and the trans-empirical in terms of logical rules is itself to beg further question. In this sense, I cannot go along with the idea to identify religious truth with the logical formula of reciprocity such as "P r)Q.Qr> P." Nevertheless, I am convinced that this kind of analysis 14

helps us to understand better as to how the same problem was approached by the ancients.

NOTES 1. The Central Conception o) Buddhism, London: 1960. 2. Kathdvatthu ed. by A.C. Taylor, PTS: 1894-7; tr. by S.Z. Aung and Mrs. Rhys Davids, PTS: 1915. 3. T h e Romanized text, ed. by Johnston and Kunst, MCB IX (1951), Appendix. sarvatra svabhdvo no vidyata iti krtvd sunydh sarvabhdvd iti/. . . . yasmdn nihsvabhdvas tasmdc chunyahl (Comment on kdrikd 1; loc. cit., p. 10) T h e pattern of argument: "whatever is devoid of oxvn-being is empty," recurs throughout the work. 4. Loc. cit., p. 52: kdrikd 70: prabhavati ca Sunyateyam yasya prabhavanti tasya sarvdrthdh/ prabhavati na tasya kirn cinna prabhavati iunyatd yasya/i 5. Sankrasvamin formulated Dignags trairupya theory in his Nydyapravesakasutram very concisely as follows: (COS Vol. 33, 1930, p. 1): paksadharmatvam sapakse sattvam vipakse cdsattvam iti// T h e process involved in anvaya and vyatireka is threefold: (1) to determine in a drstdnta the class of sapaksa, of all of whose members the two related predications are correct separately and jointly; (2) to determine in a contrapositive drstdnta the class of vipaksa, of all of whose members those two predications are not correct in the same way; and (3) to apply this concomitance of those predications to a particular class member in question. 6. Respectively, Anuloma, Patikamma, Niggaha, Upanayana, and Niggamana. 7. Madhyamikakdrikd, Chp. 2, kdrikd 10: pakso gantd gacchattti yasya tasya prasajyatel gamanena vind gantd gantur gamanam icchatahi/ 8. Vigraha., kdrikd 37 {loc. cit., p. 34): ndsti tamas ca jvalane yatra ca tisthati pardtmani jvalanah! kurute katham prakdsam sa hi prakdio 'ndhakdravadhahlI


Sign up to vote on this title
UsefulNot useful

Master Your Semester with Scribd & The New York Times

Special offer for students: Only $4.99/month.

Master Your Semester with a Special Offer from Scribd & The New York Times

Cancel anytime.